What democratic legitimacy? Analysing sample composition in the Lisbon citizens’ assembly

As echoed by international agencies, the spread of citizens’ assemblies is on the rise worldwide. Deliberation is deemed to trigger the democratisation of societies by unfolding ideal conditions for mutual learning and opinion change within randomly selected samples of citizens. Sortition draws inspiration from Ancient Greek democracy and is implemented today in different manners. In most cases, participants are invited via stratified random selection, while in other cases sortition is associated with other methods in an attempt to increase representativeness by, for example, defining quotas for the inclusion of underrepresented communities. In fact, sample composition is fundamental to understand the democratic legitimacy of assemblies’ outcomes. In Lisbon, the first citizens’ assembly ever implemented by a public authority in the country, took place in May 2022. Citizens had been encouraged to register in the city council’s website as a stratified random selection was released in April to invite 50 citizens to deliberate on the topic of climate change. At the outset, participants shared high levels of political efficacy and, while learning was ensured by sharing expert and experiential knowledge, participants showed low rates of opinion change during the session. Some degrees of homogeneity among participants and little participation from minority groups eventually convinced the city council to change the sortition method in 2023. In its second edition, a combination of registered citizens and those who will receive invitation letters from the city council will converge within the database from which sortition will be released. Acknowledging this, I will take advantage of my role as evaluator of the Lisbon citizens’ assembly to analyse and compare data retrieved from the first and second editions. I will examine in-depth the implications of the different sortition methods and discuss the change from 2022 to 2023 in terms of representativeness and inclusion. Through the lens of the different sample compositions, I will finally consider emerging issues of democratic legitimacy from Lisbon to indicate pathways for future research and debate.